Tropical Algebraic Geometry for Neuronal Representations: An Arakelov-Green Measure Based Descriptor for Graph Learning

2026-08-05

AI Summary

This work addresses the limitations of existing graph neural networks, which are constrained by the 1-Weisfeiler-Lehman (1-WL) test and thus struggle to capture ring structures induced by spatial proximity in 3D neurons and their intricate geometry-topology coupling. To overcome this, the study introduces tropical algebraic geometry and Arakelov theory into graph representation learning, proposing a training-free geometric prior method. Specifically, it employs the tropical Abel–Jacobi map to transform spatial trees into cycle-weighted metric graphs, derives node coordinates and graph-level spectral signatures via the Arakelov–Green measure, and decomposes path metrics on the universal cover of the Albanese torus using a continuous relaxation of polarized distances—thereby circumventing the NP-hard closest vector problem. The approach surpasses the expressive power of 1-WL, demonstrates strong performance on the BREC benchmark, and significantly enhances classification accuracy for VAEs, GNNs, and Tree-LSTMs on 3D neuron datasets including ACT-4, JML-4, and BIL-6, outperforming explicit lattice approximation methods.

Abstract

The quantitative analysis of 3D neuronal morphologies requires capturing both graph topology and spatial geometry. Current message-passing Graph Neural Networks (GNNs) are bounded by the 1-Weisfeiler-Lehman (1-WL) test, limiting their ability to capture cycles induced by spatial proximities. To address this, we propose a training-free geometric prior based on tropical algebraic geometry. We apply the recently established tropical Abel-Jacobi transform and polarization distances to machine learning on tree-structured data. We introduce a structural transformation pipeline, comprising cycle space augmentation and quotient space construction, to convert spatial trees into cyclic metric graphs suitable for embedding into the Tropical Jacobian. Computing exact tropical polarization distances requires solving the NP-Hard Closest Vector Problem (CVP) on integer lattices. Instead of relying on explicit approximations with quantization errors (e.g., Babai's rounding), we adopt a continuous relaxation on the universal cover of the Albanese torus. We show that the discrete Arakelov-Green measure, computed in closed form via the graph Laplacian's generalized inverse, decomposes exactly into the intrinsic path metric minus the unquantized polarization distance on this cover, avoiding integer lattice searches. This metric yields two descriptors: eigenvectors provide node-level structural coordinates, and the permutation-invariant eigenvalue spectrum provides a graph-level signature. On the BREC benchmark, the eigenvector formulation demonstrates expressivity beyond the 1-WL limit. On 3D morphology datasets (ACT-4, JML-4, BIL-6), the spectrum seamlessly integrates into standard architectures (VAEs, GNNs, Tree-LSTMs) without additional trainable parameters, outperforming explicit lattice approximations and improving classification accuracy over existing spatial models.

Problem

• Graph Neural Networks
• 1-Weisfeiler-Lehman test
• 3D neuronal morphologies
• spatial geometry
• cycle structures

Innovation

• tropical algebraic geometry
• Arakelov-Green measure
• graph neural networks
• cycle space augmentation
• polarization distance